Alma Mater Studiorum – Università di Bologna

DOTTORATO DI RICERCA IN

BIOLOGIA CELLULARE E MOLECOLARE

Ciclo XXIX

Settore Concorsuale SSC 05/E2

Settore Disciplinare SSD BIO/11

Investigating the molecular mechanisms of Neisseria meningitidis antigen regulation:

determining a switch between colonization and invasion

Presentata da: Sara Borghi

Coordinatore Dottorato

Prof. Giovanni Capranico

Relatore:

Prof. Vincenzo Scarlato

Correlatore:

Dott.ssa Isabel Delany

Esame finale anno 2017

2

Table of Contents

1. Abstract ........................................................................................................................................ 4

2. Introduction ................................................................................................................................. 6

2.1 Meningococcal disease ......................................................................................................... 6

2.2 Neisseria meningitidis: pathogen and pathogenesis ........................................................... 7

2.3 Meningococcal virulence factors ....................................................................................... 11

2.4 Meningococcal vaccines ..................................................................................................... 17

2.4.1 Licensed vaccines against MenB ................................................................................. 18

2.4.2 Investigational MenB vaccines .................................................................................... 20

Chapter 1: NHBA regulation and expression during colonization and invasion

Sensing the environment .......................................................................................................................... 24

Neisserial Heparin Binding Antigen (NHBA) ...................................................................................... 25

3. Results ........................................................................................................................................ 28

3.1 NHBA expression and surface exposure are temperature-dependent ........................ 28

3.2 Mutations and deletions in the 5’UTR and 5’TR of nhba affect expression ................ 31

3.3 NHBA is expressed during the active growth ................................................................ 38

3.4 NHBA thermoregulation is not driven by the nhba promoter ...................................... 38

3.5 Temperature affetcs nhba RNA half-life ........................................................................... 41

3.6 NHBA protein shows higher stability at 30°C respect to 37°C ..................................... 44

3.7 NHBA expression levels correlate with susceptibility to complement-mediated killing by anti-NHBA antibodies ............................................................................................... 45

3.8 NHBA regulation during invasion ................................................................................... 50

3.9 NHBA cleavage does not affect NHBA-mediated killing suceptibility ...................... 54

4. Discussion and conclusions ................................................................................................... 56

Chapter 2: Comparing different delivery systems by using NadA as a model antigen

Immune response and vaccine design ................................................................................................... 67

Neisserial adhesin A (NadA) ................................................................................................................... 69

5. Results ........................................................................................................................................ 72

5.1 NadA overexpression on MenB and E.coli nOMV ......................................................... 72

5.2 Prototype nOMV vaccines elicited higher titre of α-NadA antibodies ....................... 74

5.3 Evaluation of the bactericidal activity of the antibodies elicited through rSBA assay ............................................................................................................................................... 76

5.4 Inhibition of E.coli-NadA var3 adhesion to Chang epithelial cells with different sera .................................................................................................................................................. 79

3

6. Discussion and conclusions ................................................................................................... 81

7. Materials and methods ............................................................................................................ 86

7.1 Bacterial strains and culture conditions ........................................................................... 86

7.2 Generation of plasmids and N.meningitidis new recombinant strains ....................... 86

7.3 Polyacrylamide gel electrophoresis and Western blotting ............................................ 88

7.4 Quantitative real-time PCR (qRT-PCR) experiments ..................................................... 88

7.5 Flow cytometry .................................................................................................................... 89

7.6 Serum Bactericidal Assay (SBA) ........................................................................................ 90

7.7 RNA stability assay ............................................................................................................. 90

7.8 Protein stability assay ......................................................................................................... 90

7.9 nOMV preparation .............................................................................................................. 91

7.10 Transmission Electron Microscopy ................................................................................. 92

7.11 Mice immunizations ......................................................................................................... 93

7.12 IgG antibody titers elicited against NadA ..................................................................... 93

7.13 Inhibition of the binding assay ........................................................................................ 94

8. Appendix .................................................................................................................................... 95

9. References ................................................................................................................................ 104

4

1. Abstract

Neisseria meningitidis colonizes the nasopharynx of humans and pathogenic strains can

disseminate into the bloodstream causing septicemia and meningitis. Neisserial

Heparin Binding Antigen (NHBA) and Neisserial Adhesin A (NadA) are part of a

multicomponent vaccine against N. meningitidis serogroup B, called 4CMenB

(Bexsero™).

NHBA is a surface-exposed lipoprotein which is expressed by all N. meningitidis strains

in different isoforms. NHBA harbors an arginine-rich motif through which it is able to

bind heparin-like molecules, increasing adherence to host tissues and heparin-

mediated serum resistance. We determined that temperature controlled the expression

of NHBA in all strains tested, regardless of the clonal complex or peptide isoform

expressed. NHBA expression was significantly increased at 30-32°C compared to 37°C,

the temperature standardly used for in vitro culturing. An increase in NHBA

expression at lower temperatures was measurable both at protein and RNA levels and

was also reflected by a higher surface exposure of this antigen. A detailed molecular

analysis indicated that multiple molecular mechanisms are responsible for the

thermoregulated NHBA expression. The comparison of RNA steady state levels in cells

cultured at 30°C and 37°C demonstrated an increased RNA stability/translatability at

lower temperatures. Furthermore, protein stability was also impacted resulting in

higher NHBA stability at lower temperatures. Increased NHBA expression resulted in

more efficient killing as shown by serum bactericidal assay (SBA). Mimicking the

invasive condition, we investigated the NHBA expression in response to the presence

of serum. We showed that the presence of human serum has contrasting effects on

NHBA expression, resulting in transient up-regulation of NHBA at transcriptional

level, however the protein is rapidly processed likely by complement proteases. We

propose a model in which NHBA regulation in response to temperature downshift

might reflect the bacterial adaptation during the initial step of host-bacterial interaction

and might also explain higher susceptibility to anti-NHBA antibodies in the

nasopharynx niche. On the other hand, the initial up-regulation and the high

processing of NHBA might play a role during the first steps of invasive disease. All

together these data describe the importance of NHBA both as virulence factor and as

vaccine antigen during neisserial colonization and invasion.

5

In the second part of the thesis, we compared genetically engineered outer membrane

vesicles and recombinant proteins, as delivery systems of protective antigen. Using

NadA as model antigen, we determined that OMV overexpressing NadA produced by

homologous (MenB) or heterologous (E.coli) bacterial strains, are able to elicit a higher

functional antibody response respect to the recombinant protein per se, despite the

comparable anti-NadA titres elicited. The differences in functionality might be due to

different IgG subclasses distribution. Moreover, OMV overexpressing NadA are able to

elicit antibodies that inhibit NadA-mediated adhesion on the host cells surface, in a

much more efficient way respect to the recombinant protein formulation.

These results indicate that the antigen delivered on OMV triggers a good functional

immune response. This preliminary characterization supports the use of OMV as

delivery systems for next generation vaccine design and remark the great potential of

NadA as model antigen.

6

2. Introduction

2.1. Meningococcal disease

Invasive meningococcal disease is characterized by a rapid onset and progression to

meningitis and/or sepsis which can lead to death within hours. The etiological agent of

this devastating disease is Neisseria meningitidis, otherwise known as meningococcus.

The first report of meningococcal disease is dated back to 1887 by Anton

Weichselbaum, who described the meningococcal infection of the cerebrospinal fluid of

a patient (Weichselbaum, A. 1887). Each year there are an estimated 1.2 million cases of

invasive meningococcal disease and 135,000 deaths (WHO 2010). Despite the

availability of antibiotic treatment, approximately 10 to 14% of people who contract

meningococcal disease die, and the rate increases to 40-55% in the case of sepsis

(Brandtzaeg, P. et al. 2005, Rosenstein, N.E. et al. 2001). Furthermore, approximately 11

to 19% of individuals surviving the disease often suffer from permanent sequelae,

including neuro-developmental deficits, hearing loss, ataxia, hemiplegia, seizures and

limbs loss (Kaplan, S.L. et al. 2006, Rosenstein, N.E. et al. 2001, Thompson, M.J. et al.

2006, WHO 2010).

Multiple studies have demonstrated that carriage rates are very low in the first few

years of life, but rise during adolescence, reaching peaks of 10-35%, and decreasing to

less than 10% in older groups (Caugant, D.A. et al. 2007, Claus, H. et al. 2005). In

contrast to carriage rates, meningococcal disease is rare, varying from 0.5 to 10 per

100,000 persons; however, the incidence can rise above 1 per 1,000 persons during

epidemics (Caugant, D.A. et al. 2009, Stephens, D.S. et al. 2007). Most cases of

meningococcal disease occur in otherwise healthy individuals without identified risk

factors and what determines the transition from colonization to invasive disease is not

yet fully understood. However, certain biological, environmental and social factors

have been associated with an increased risk of disease. Infants under one year of age

have the highest risk of infection due to their immature immune systems (6.33-7.08

cases per 100,000). Whereas, the peak observed in adolescents is largely due to

increased carriage in this population (Cohn, A.C. et al. 2010). Several studies

demonstrated that both host and pathogen factors influence the development of the

disease, such as human genetic polymorphisms, impaired immune system, microbial

virulence factors, as well as environmental conditions facilitating exposure and

7

acquisition, and naso- and oro-pharyngeal irritation caused by smoking and

respiratory tract infection (Brigham, K.S. et al. 2009, Davila, S. et al. 2010,

Goldschneider, I. et al. 1969, Harrison, L.H. 2006, Imrey, P.B. et al. 1995, Rosenstein,

N.E. et al. 2001, Zuschneid, I. et al. 2008). The unspecific symptoms at the onset and the

early stage of infection, like headache, fever and rash, can implicate an arduous

diagnosis. Due to the rapid progression of this life-threatening pathology, vaccination

represents the unique effective public health response.

2.2. Neisseria meningitidis: pathogen and pathogenesis

N. meningitidis is a strictly human, Gram-negative β-proteobacterium member of the

Neisseriaceae family. It is an aerobic, non-motile and non-sporulating diplococcus

(Figure 2.1), usually encapsulated and piliated.

Figure 2.1. Immuno-gold labelling and transmission electron microscopy of Neisseria

meningitidis. Analysis of the strain was performed with antisera raised against the NadA

adhesin. Scale bars: 200 nm (from (Pizza, M. et al. 2000)

The envelope of N. meningitidis consists of the cytoplasmic membrane, the outer

membrane (OM) and the periplasm between them, which contains a layer of

peptidoglycan. The cytoplasmic membrane is a phospholipid bilayer, whereas the OM

is composed of a phospholipidic inner leaflet and an outer leaflet of

lipooligosaccharide (LOS). Some meningococcal strains have a polysaccharide capsule

8

attached to their OM and almost all pathogenic strains are encapsulated. Nevertheless,

also non-encapsulated isolates have been recently associated to invasive disease

(Johswich, K.O. et al. 2012). On the basis of the bacterial polysaccharide capsule, N.

meningitidis can be classified into at least thirteen serogroups: A, B, C, E-29, H, I, K, L,

W, X, Y, Z and 29E (Branham, S.E. 1953). Among them, six serogroups (A, B, C, X, Y

and W) are responsible for more than 90% of meningococcal disease worldwide and

are thus considered pathogenic (Boisier, P. et al. 2007, Frasch, C.E. 1989, Jarvis, G.A. et

al. 1987, Stephens, D.S. et al. 2007). Meningococci are further classified into serotypes

and serosubtypes according to antigenic differences in their major outer membrane

proteins, PorA and PorB. However, the classification based on the serological

characteristics of N. meningitidis is limited due to the high frequency of variation of

OM-proteins, probably determined by a strong selective pressure. Hence, new DNA-

based methods for the characterization of meningococcal isolates have been developed,

and the Multi Locus Sequence Typing (MLST) is now considered the gold standard for

molecular typing and epidemiologic studies (Maiden, M.C. et al. 1998). This typing

system relies on polymorphisms within seven housekeeping genes; each sequence for a

given locus is screened for identity with already known sequences for that locus. If the

sequence is different, it is considered to be a new allele and an identification number is

assigned. Therefore, the combination of the seven allele numbers determines the allelic

profile of the strain, and each different allelic profile is assigned as a sequence type

(ST). Meningococci sharing at least four of the seven loci with a central ancestral

genotype are grouped together into clonal complexes (CCs) (Urwin, R. et al. 2003).

Through the employment of MLST it has been shown that the majority of strains

associated with invasive disease belong to specific CCs (ST-1, ST-4, ST-5, ST-8, ST-11,

ST-32, ST41/44 and ST-269), called hyper-invasive (Caugant, D.A. 2008, Maiden, M.C.

2008). However, the reasons of this enhanced pathogenic phenotype are yet unknown.

9

The pathogenesis of N. meningitidis is a complex multi-stage process (Figure 2.2).

Figure 2.2 Stages in the pathogenesis of Neisseria meningitidis. N. meningitidis may be

acquired through the inhalation of respiratory droplets. The organism establishes intimate

contact with non-ciliated mucosal epithelial cells of the upper respiratory tract, where it may

enter the cells briefly before migrating back to the apical surfaces of the cells for transmission to

a new host. Besides transcytosis, N. meningitidis can cross the epithelium either directly

following damage to the monolayer integrity or through phagocytes in a ‘Trojan horse’ manner.

In susceptible individuals, once inside the blood, N. meningitidis may survive, multiply rapidly

and disseminate throughout the body and the brain. Meningococcal passage across the brain

vascular endothelium (or the epithelium of the choroid plexus) may then occur, resulting in

infection of the meninges and the cerebrospinal fluid (See text for details) (Virji, M. 2009).

The human nasopharynx is the natural biological niche colonized by N. meningitidis

and transmission to new hosts is facilitated through aerosol droplets (Caugant, D.A.

and Maiden, M.C. 2009), as well as direct contact. Acquisition is generally

asymptomatic, but infrequently may result in local inflammation, invasion of mucosal

surfaces, access to the bloodstream and fulminant sepsis or focal infections such as

meningitis (Stephens, D.S. et al.). Meningococcal disease usually occurs 1–14 days after

acquisition of the pathogen (Rosenstein, N.E. et al. 2001), after which the carrier state

may be established for a period that vary between days to months. From an

evolutionary perspective, the interactions of meningococci and the human

nasopharynx are key events. Meningococcal carriage and transmission, not disease,

10

determine the global variation and composition of the natural population of

meningococci.

Colonization is an essential but considerably challenging process in meningococcal

survival, and therefore a prerequisite for strain carriage as well as for establishing

invasive disease (Stephens, D.S.).

Initially, N. meningitidis preferentially adheres to relatively non-ciliated or damaged

areas of the epithelial barrier. Pili and outer membrane opacity proteins such as Opa

and Opc play a major role for meningococcal adhesion to human cells (Hill, D.J. et al.

2012, Kallstrom, H. et al. 2001). Upon contact with human cells, the meningococcus

forms microcolonies and adheres using filamentous structures named type IV pili

(T4P) which may recognize the host receptor CD46 (Kallstrom, H. et al. 2001), forming

a layer tightly attached to host cells (Nassif, X. et al. 1997). After this step, the capsule,

which masks the OM proteins via steric hindrance, is lost or down-regulated due to

cell-contact induced repression (Deghmane, A.E. et al. 2002) or selection of low or no-

capsule expressing bacteria caused by phase variation (Hammerschmidt, S. et al. 1996).

The absence of the capsule reveals a variety of redundant adhesins, which mediate a

close adherence of the bacteria to host epithelial cells (Stephens, D.S. 2009). In fact,

other minor adhesins such as Neisseria adhesin A (NadA) (Capecchi, B. et al. 2005),

Neisseria hia/hsf homologue (NhhA) (Scarselli, M. et al. 2006), Adhesin complex

protein (Acp) (Hung, M.C. et al. 2013), Adhesion and penetration protein (App)

(Serruto, D. et al. 2003), Meningococcal serine protease A (MspA) (Turner, D.P. et al.

2006) and Neisserial heparin binding antingen (NHBA) (Vacca, I. et al. 2016) have been

shown to significantly contribute towards N. meningitidis colonization of the human

nasopharynx.

The interaction of bacterial opacity proteins, Opa and Opc, with CD66/CEACAMs and

integrins respectively, on the surface of epithelial cells triggers meningococcal

internalization (Gray-Owen, S.D. et al. 2006). Meningococci are capable of intracellular

replication and this is in part due to iron acquisition mediated by specialized transport

systems, such as the transferring binding protein (TbpAB), the lactoferrin binding

protein (LbpAB), and the hemoglobin binding receptor (HmbR) (Perkins-Balding, D. et

al. 2004). This intracellular lifestyle gives the bacteria the opportunity to evade the host

immune response as well as to find new source of nutrients. Occasionally, bacteria can

cross the mucosal epithelial barrier of susceptible individuals, either through

11

transcytosis or through phagocytes in a “Trojan horse” manner, or directly following

damage to the monolayer integrity (Virji, M. 2009), and eventually enter the

bloodstream. In healthy individuals, bacteria that cross the mucosal epithelium are

eliminated by serum bactericidal activity. Nonetheless, survival within human blood

relies upon different mechanisms and is dependent on their capability to evade the

immune response and to acquire nutrients. Indeed, the up-regulation of capsule

expression prevents antibodies and complement deposition (Uria, M.J. et al. 2008)

hence inhibiting phagocytosis. Other strategies developed by the bacteria to evade the

immune system are the recruitment of negative regulators of the complement cascades,

such as Factor H (fH), which is bound by the Factor H binding protein (fHbp) (Madico,

G. et al. 2006), or by the Neisserial surface protein A (NspA) (Lewis, L.A. et al. 2010),

and by the Porin B (PorB) (Lewis, L.A. et al. 2013), or the recruitment of complement

regulators, such as the C4-binding protein, which is bound by Porin A (PorA) (Jarva,

H. et al. 2005). Once inside the bloodstream, meningococci can multiply slowly and

eventually cross the blood-brain barrier, causing the infection of meninges and

cerebrospinal fluid (Nassif, X. 2009). Otherwise, in case of rapid multiplication within

the blood, the bacteria cause septicemia or meningococcemia (Rosenstein, N.E. et al.

2001, Tinsley, C. et al. 2001).

Overall, the onset of meningococcal disease can be seen as a failed relationship

between the meningococcus and the host. While factors that trigger meningococcal

entrance in the bloodstream are not yet fully understood, they are likely dependent on

both the host and pathogen sides and include impairing of the integrity of the human

nasopharyngeal mucosa, the lack of a protective immune response and microbial

factors influencing virulence (Caugant, D.A. and Maiden, M.C. 2009, Stephens, D.S. et

al. 2007).

2.3. Meningococcal virulence factors

Within the host N. meningitidis colonizes and invades diverse sites which represent

different niches with respect to nutrients, environmental factors and competing

microorganisms. The pathogen is subjected to constant selective pressures and its

ability to rapidly adapt its metabolism and cellular composition to environmental

changes is essential for its survival (Hill, D.J. et al. 2010). Bacteria achieve adaptation to

the environment either by changing their genotype (genome plasticity) or by transient

12

alterations in gene expression. These two mechanisms are complementary and both

lead to phenotypic variations.

The genome variability is also assured by the horizontal gene transfer that occurs with

a relatively high frequency considering the high natural competence of meningococci.

Instead, the genome plasticity is guaranteed by the abundance of mobile elements that

represent the 10% of the entire genome (Parkhill, J. et al. 2000). Furthermore, other

interesting phenomenon that significantly contributes to meningococcal genome

plasticity is phase variation. It represents an adaptive process by which N. meningitidis

undergoes stochastic, frequent and reversible phenotypic changes as consequence of

genetic alterations in specific loci, altering mainly virulence-associated, surface-

exposed antigens such as outer-membrane proteins PorA, Opc, Opa, pili and specific

adhesins, as well as LOS and capsule (Davidsen, T. et al. 2006, Feil, E.J. et al. 2001,

Metruccio, M.M. et al. 2009, Moxon, E.R. et al. 1998). Meningococcal strains associated

with disease have high frequency of phase variation, indicating that varying surface-

exposed components provides substantial benefits during transmission between hosts

(Richardson, A.R. et al. 2002).

Distinct from phase variation, antigenic variation is a mechanism of immune evasion

where bacteria express different moieties of functionally conserved molecules that are

antigenically distinct within a clonal population. This process is distinct from phase

variation, as only one variant is expressed at any given time, although the cell still

contains the genetic information to produce a whole range of antigenic variants. In the

pathogenic Neisseria species, antigenic variation occurs in several surface components,

including type IV pili, LOS and Opa proteins (Davidsen, T. and Tonjum, T. 2006).

The virulence of N. meningitidis is influenced by multiple factors that are mainly

located in the outer membrane (Figure 2.3).

13

Figure 1.3 Meningococcal cell compartments. Schematic representation of the different

bacterial compartments and of the main components of the outer membrane, together with their

known function (adapted from Rosenstein NE, 2001).

The main virulence factor is the polysaccharidic capsule, which represents a barrier

that protects the bacterium from the host innate and adaptive immune system

(Schneider, Exley, Ram, Sim, & Tang, 2007; Vogel & Frosch, 1999). It also defends

meningococcus from desiccation during airborne transmission between hosts (Virji, M.

2009); (Romero, J.D. et al. 1997). Its expression is phase variable (Hammerschmidt, S. et

al. 1996) and the switching of the capsule locus between strains confers a selective

advantage to the bacterium for its evasion to opsonization or neutralization by natural

or vaccine-induced anti-capsular antibodies (Swartley, J.S. et al. 1997). In

meningococcus, LPS are referred to as lipooligosaccharide (LOS) because of the

presence of repeating short saccharides instead of long-chain saccharides. LOS is the

major constituent of the outer leaflet of the meningococcal outer membrane (OM),

responsible for the physical integrity and proper functioning of the membrane and

required for resistance of N. meningitidis to complement (Geoffroy, M.C. et al. 2003).

LOS comprises an inner and outer oligosaccharide core attached to the lipid A portion

that anchors the LOS in the outer leaflet of the OM. Lipid A is responsible for the

toxicity of LOS due to its ability to bind to different Toll-like receptors on monocytes

and on dendritic cells triggering the secretion of various inflammation mediators

(Brandtzaeg, P. et al. 2001); (Braun, J.M. et al. 2002). Phase and antigenic variations lead

to different saccharide chains altering dramatically the antigenic properties of LOS and

enabling individual meningococci to display a repertoire of multiple LOS structures

14

simultaneously (Jennings, M.P. et al. 1999). Another group of virulence factors

involved in the interface between the meningococcus and the host are pili. They are

long filamentous structures consisting of protein subunits that extend from the

bacterial surface beyond the capsule (Pinner, R.W. et al. 1991, Virji, M. et al. 1992). Pili

represent the major contributor to the adhesive property of the capsule (Stephens, D.S.

et al. 1981, Virji, M. et al. 1991) and in addition they are involved in the uptake of

foreign DNA from the extracellular environment, hence increasing transformation

frequency and consequently genetic adaptability (Helaine, S. et al. 2007).

Furthermore, the presence of other OM-associated proteins is important in host cell

interaction. The opacity proteins (Opa and Opc) are integral outer membrane proteins

that mediate pathogen-host interaction, adhering to and invading of epithelial and

endothelial cells (Virji, M. et al. 1992). A key role in the adhesion is carried out by

adhesins, which are generally low expressed in vitro, but they might be upregulated in

vivo. In fact, they may undergo to antigenic variation and/or phase variation, hence

allowing the meningococcus to evade the immune system and adapt to different niches

(Virji, M. 2009). The Neisserial adhesin A (NadA) is a surface-exposed member of the

Oligomeric coiled-coil adhesin family of bacterial Trimeric Autotransporter adhesins

(El Tahir, Y. et al. 2001, Helaine, S. et al. 2007). NadA mediates adhesion to and

invasion of human epithelial cells (Capecchi, B. et al. 2005), suggesting its pivotal role

in the adhesion to the naso- and oro-pharyngeal epithelia during meningococcal

colonization of the human upper respiratory tract. Other adhesins have been reported

to play a role in colonization and/or invasion. NHBA has been recently shown to

participate during the colonization process by increasing adherence to host tissues by

binding glycosaminoglycans (Vacca, I. et al. 2016), and mediating biofilm formation

((Arenas, J. et al. 2013)). The Meningococcal surface fibril (Msf), previously termed

Neisseria hia/hsf homologue A (NhhA) (Peak, I.R. et al. 2000, Weynants, V.E. et al.

2007), mediates adhesion to epithelial cells and to components of the extracellular

matrix, even though at low levels (Scarselli, M. et al. 2006). Moreover, it has been

shown its involvement in the immune system evasion. Msf binds to the activated form

of Vitronectin and inhibits the terminal complement pathway (Griffiths, N.J. et al.

2011), and its role in inhibiting phagocytosis, inducing macrophages apoptosis and

protecting bacteria against complement-mediated killing has been suggested

(Sjolinder, H. et al. 2008, Sjolinder, M. et al. 2012). Two homologous autotransporters,

15

the Adhesion penetration protein (App) and the Meningococcal serine protease A

(MspA) are involved in the bacterial interaction to epithelial cells (Serruto, D. et al.

2003, Turner, D.P. et al. 2006) and also in the apoptosis of dendritic cells (Khairalla, A.S.

et al. 2015). Glycolipid adhesins such as members of the Multiple adhesin family (Maf)

may contribute to the bacterial interaction with host cells (van Putten, J.P. et al. 1998).

Interestingly they are found to be associated with genomic islands present only in

pathogenic Neisseria species, both meningococcus and gonococcus (Jamet, A. et al.

2015).

The two porins PorA and PorB, are the most abundant proteins present in the

meningococcal OM. They are composed of relatively conserved regions, which are

predicted to form the β-barrel structure that spans through the membrane, alternated

with variable regions, which should be surface-exposed, hence undergoing to a strong

selective pressure. The formation of trimers creates the pore structure that allows the

passage of small hydrophilic solutes necessary for the bacterial metabolism. Porins

were shown to be interacting with several human cell types and proteins (Orihuela,

C.J. et al. 2009); moreover, PorA elicits a protective immune response in humans

(Holst, J. et al. 2009, Wedege, E. et al. 1998), while PorB might be involved in the

immune system evasion by binding the human fH (hfH) (Lewis, L.A. et al. 2013). The

regions of PorA that generate the immune response are loops 1 and 4, named VR1 and

VR2, that are hyper variable among strains. OMV based vaccines, such as 4CMenB, use

PorA as significant antigen generating bactericidal immune responses. However, due

to the hyper variability of the immune-dominant regions, PorA-based vaccines provide

protection only against strains expressing homologous PorA serosubtypes (see below).

Furthermore, the genome of N. meningitidis contains a set of membrane-associated

factors responsible for the host’s immune system evasion and hence for its virulence.

As indicated by the elevated susceptibility to microbial, including meningococcal,

infections exhibited by individuals with complement deficiencies (Figueroa, J. et al.

1993). In order to escape from the innate immune system, N. meningitidis has evolved a

plethora of mechanisms that target the complement cascades. As already introduced

above, at least three meningococcal proteins have shown to bind the fH, fHbp (Madico,

G. et al. 2006), NspA (Lewis, L.A. et al. 2010) and PorB (Lewis, L.A. et al. 2013). Strains

lacking both fHbp and NspA were not able to bind fH and indeed were more

susceptible to complement-dependent killing (Echenique-Rivera, H. et al. 2011, Lewis,

16

L.A. et al. 2010). In addition, the observed binding of heparin from the NHBA may

increase bacterial serum resistance due to the potential interactions of heparin with fH

(Serruto, D. et al. 2010).

17

2.4. Anti-meningococcal vaccines

Due to its rapid progression and the difficulties to diagnose it (Rosenstein, N.E. et al.

2001, Thompson, M.J. et al. 2006), the most effective option to prevent meningococcal

disease is vaccination. No broadly protective vaccine is currently available to provide

protection against all serogroups of N. meningitidis. Different meningococcal vaccines

have been developed against the distinct serogroups (Zahlanie, Y.C. et al. 2014). There

are a number of polysaccharide and conjugate meningococcal vaccines in use against

serogroups A, C, Y and W135. The tetravalent vaccine composed of purified capsular

polysaccharides, although efficacious in adolescent and adults, is poorly immunogenic

in infants and fails to induce immunological memory. However, when conjugated to a

carrier protein, capsule polysaccharides show a greatly improved immunogenicity in

young infants (Granoff, D.M. et al. 2007, Nassif, X. 2009, Virji, M. 2009). Monovalent,

bivalent, and tetravalent polysaccharide conjugative vaccines are available and

effective against meningococcal serogroups A, C, Y and W-135 (Zahlanie, Y.C. et al.

2014); http://www.who.int/ith/vaccines/meningococcal/en/]. The first trials

conducted in the United Kingdom with the meningococcus C conjugate showed a

dramatic decline in the incidence of serogroup C disease in all age groups that received

the vaccine (Borrow, R. et al. 2000, Miller, E. et al. 2001) with an efficacy of 97 and 92

per cent for teenagers and toddlers, respectively (Ramsay, M.E. et al. 2001).

In contrast, the group B capsule polysaccharide is not suitable as vaccine antigen. It

consists of a homolinear polymer of α(2→8)N-acetyl neuraminic acid, also known as

polysialic acid, which is structurally similar to the sialic acid found in human neural

tissue, hence is poorly immunogenic in humans and may elicit auto-antibodies (Finne,

J. et al. 1987, Finne, J. et al. 1983). Therefore, efforts to develop a vaccine against

meningococcus serogroup B (MenB) focused mainly on non-capsular antigens, such as

proteins or LOS. The principal challenge has been to identify surface-exposed non-

capsular antigens that are safe, antigenically conserved and that elicit a broad Serum

Bactericidal Antibody (SBA) response. Licensed and promising group B vaccine

approaches are discussed below.

http://www.who.int/ith/vaccines/meningococcal/en/

18

2.4.1. Licensed vaccines against MenB

Detergent-extracted OMV vaccines (dOMV)

In order to control outbreaks caused by specific MenB strains vaccines composed of

dOMV have been successfully employed in Norway (Fredriksen, J.H. et al. 1991), Cuba

(Sierra, G.V. et al. 1991), Chile (Boslego, J. et al. 1995) and New Zealand (Oster, P. et al.

2005). The detergent treatment removes the toxic LOS, but it also extracts other

desirable antigens such as lipoproteins. Consequently, the porin protein PorA results

to be the immuno-dominant antigen (Martin, D.R. et al. 2006, Tappero, J.W. et al. 1999).

Despite dOMV vaccines resulted to be safe and to induce good functional responses in

humans, the immune response elicited is effective only against strains expressing the

same PorA serosubtype, due to PorA antigenic variability (van der Ley, P. et al. 1991).

Therefore, dOMV vaccines are well-suited to control local, clonal outbreaks but they do

not confer broad protection.

4CMenB

The advent of the genomic era and the availability of whole genome sequences have

contributed to radically change the approach to vaccine development. Indeed, the in

silico approach named Reverse Vaccinology (RV) aims to identify surface-exposed

non-capsular antigens that are antigenically conserved among strains and elicit a

bactericidal serum response. This approach led to the development of the four

component recombinant protein vaccine 4CMenB (Giuliani, M.M. et al. 2006, Giuliani,

M.M. et al. 2010). 4CMenB contains five Genome-derived Neisseria Antigens (GNA)

formulated together with the dOMV component from the NZ98/254 strain (Martin,

D.R. et al. 2006). Based on their ability to induce broad protection three major antigens

have been selected (Giuliani, M.M. et al. 2006): NadA (Capecchi, B. et al. 2005,

Comanducci, M. et al. 2002) is present as single polypeptide, while fHbp (Beernink,

P.T. et al. 2008, Masignani, V. et al. 2003) and NHBA (Serruto, D. et al. 2010, Welsch,

J.A. et al. 2003) are fused to the conserved meningococcal gene products GNA2091 and

GNA1030, respectively. The other two antigens, GNA2091 and GNA1030, are well

conserved in N. meningitidis, but less functionally characterized than the other antigens

(Bos, M.P. et al. 2014, Donnarumma, D. et al. 2015, Muzzi, A. et al. 2013). They were

19

included in the vaccine formulation since they increase immune responses to the main

vaccine antigens when present as fusion proteins with the respect of the individual

antigens (Giuliani, M.M. et al. 2006). 4CMenB was licensed in Europe in 2013 and in the

U.S. in 2015, following its progression through clinical trials that have demonstrated its

safety (Esposito, S. et al. 2014, Prymula, R. et al. 2014, Toneatto, D. et al. 2011) and its

efficacy in inducing a protective immune response in infants, children, adolescents and

adults against the majority of MenB strains (Gossger, N. et al. 2012, Kimura, A. et al.

2011, McQuaid, F. et al. 2014, Read, R.C. et al. 2014, Santolaya, M.E. et al. 2012, Snape,

M.D. et al. 2013, Vesikari, T. et al. 2013).

Figure 2.4 Schematic representation of the 4CMenB vaccine antigens on the surface

of N. meningitidis (from Serruto D, 2012). The different bacterial compartments (outer

membrane, periplasmic space, cytoplasmic membrane) and the main antigens

identified through reverse vaccinology approach (NHBA, fHbp and NadA) are

depicted. Other components of the meningococcal membranes are also shown (pilus,

polysaccharide capsule, lipooligosaccharide and integral inner and outer membrane

proteins).

Bivalent fHbp-based vaccine or Trumenba

Trumenba was licensed in the U.S. in 2014 for a target population of adolescents and

young adults. However, it is not suitable for use in infants considering that it consists

of purified lipoproteins known as TLR-2 agonists (Richmond, P.C. et al. 2012). It is a

recombinant protein-based vaccine composed of equal amounts of two variants,

20

subfamily A05/var3.45 and subfamily B01/var1.55, of lipidated fHbp (Fletcher, L.D. et

al. 2004).

2.4.2. Investigational MenB vaccines

Recombinant protein vaccines

Several protein antigens have been investigated for their protective ability for use in a

MenB vaccine, among which NspA, TbpB, FetA, ZnuD and others (Halperin, S.A. et al.

2007, Hubert, K. et al. 2013, West, D. et al. 2001). The main issue with all of these

approaches was the limited cross protective potential of any one antigen. It was clear

that a multivalent approach was needed to guarantee a wide protection.

Native Outer Membrane Vesicles (nOMV) vaccines

nOMV are spherical portions of the OM, ~20–250 nm in diameter, produced by Gram

negative bacteria. They are spontaneously released during the active growth into the

surrounding medium. These portions of the OM bud and detach from the cell,

enclosing many native bacterial antigens and periplasmic constituents (Figure 2.5).

The vesicles play diverse roles like delivery of virulence factors , modulation of the

host immune system during pathogenesis, aid in nutrient acquisition, mediation of

cellular communication, surface modifications and the elimination of undesired

components that, ultimately, make them a transportable part of the bacterial arsenal

and survival system (Collins, B.S. 2011, Kuehn, M.J. et al. 2005, Schwechheimer, C. et

al. 2015).

21

Figure 2.5 Model of native Outer Membrane Vesicles (nOMV) biogenesis. NOMV vesicles are

proteoliposomes consisting of OM phospholipids and LPS, a subset of OM proteins and

periplasmic (luminal) proteins (Kuehn, M.J. and Kesty, N.C. 2005).

nOMV represent an attractive vaccine platform mimicking the bacterial cell surface.

Since nOMV do not undergo to a detergent extraction procedure, like dOMV do, they

preserve high amounts of lipooligosaccharide (LOS) as well as protective lipoproteins

which would otherwise be removed by the detergent. This was expected to improve

immunogenicity and cross-protection provided but it raised safety concerns.

Consequently, to prepare safe nOMV vaccines the strain must be genetically

engineered to reduce the LOS reactogenicity. The acylation of lipid A molecule is

responsible for its endotoxin activity and two mutations (lpxL1 and lpxL2), affecting

the reduction of lipid A, have been successfully exploited (Bonvehi, P. et al. 2010,

Keiser, P.B. et al. 2011, Keiser, P.B. et al. 2010, Koeberling, O. et al. 2011). The lpxL1

gene, homologous to E.coli htrB, encodes for a late acyltransferase of lipid A

biosynthesis. Its deletion lead to penta- instead hexa-acetylated molecules, resulting in

lower endotoxin activity of LOS (van der Ley, P. et al. 2001). Instead the lpxL2 gene,

homologous of E.coli lpxLM, encodes for a lauroyl acyltransferase. Its deletion leads to

a tetra-acylated lipid A lacking both secondary lauroyl chains.

nOMV vaccines prepared from wild-type strain were poorly immunogenic in mice

(Koeberling, O. et al. 2008, Moe, G.R. et al. 2002). Koeberling and colleagues

22

demonstrated that the level of expression of a key antigen, as fHbp, was a critical

parameter to elicit broad serum bactericidal responses (Koeberling, O. et al. 2011). The

overexpression of some antigens (Keiser, P.B. et al. 2011) and the removal of the

immunodominant PorA antigen (Bonvehi, P. et al. 2010) were two strategies tested in a

phase I clinical trial. In the first case, the nOMV vaccine resulted to be safe and

immunogenic (Keiser, P.B. et al. 2011); nevertheless, the major contribution to

bactericidal activities was from antibodies raised from LOS providing immunotype

specific bactericidal responses. This result was suggested to be due to insufficient

levels of the antigens over-expressed on the vesicles (Koeberling, O. et al. 2011). In the

second case, PorA was deleted to avoid its immune-dominance. The prototype vaccine

strain was also engineered to express a truncated form of LOS immunotype L3,7 that is

the most common in invasive MenB strain (Scholten, R.J. et al. 1994). This OMV-based

vaccine offered good safety but low immunogenicity in healthy young adults (Bonvehi,

P. et al. 2010, Weynants, V. et al. 2009).

23

Chapter 1

NHBA regulation and expression during colonization and invasion

24

Sensing the environment

Colonization is an essential as well as a considerably challenging process in

meningococcal survival, and therefore a prerequisite for strain carriage as well as for

establishing invasive disease (Stephens, D.S. 2009). The nasopharyngeal epithelium is a

complex ecological niche that poses several hurdles for bacterial colonization and

survival. Compounds such as mucus, antimicrobial peptides and immunoglobulins

provide physical and biochemical host defenses (Laver, J.R. et al. 2015). Furthermore,

this environment is deprived of nutrients such as iron, carbohydrates and oxygen

essential for bacterial growth and N. meningitidis therefore needs to compete for these

limited nutrients with the resident microflora. Taken together, these factors make de

novo colonization and survival challenging. N. meningitidis is incredibly well adapted

to this environment and has developed several mechanisms to control expression of

adhesion molecules (Deghmane, A.E. et al. 2002) (Grifantini, R. et al. 2002, Hey, A. et al.

2013), biofilm formation (Arenas, J. et al. 2013), iron acquisition (Larson, J.A. et al. 2002,

Schryvers, A.B. et al. 1999), metabolism (Jamet, A. et al. 2009, Mendum, T.A. et al. 2011)

(Laver, J.R. et al. 2015), and immune evasion factors (Lomholt, H. et al. 1992,

Yazdankhah, S.P. et al. 2004) .

One of the key signals sensed by N. meningitidis to determine its environment and to

induce the expression of either adhesion or immune evasion factors is temperature

(Laver, J.R. et al. 2015). Temperatures within the upper respiratory tract are affected by

the passage of air during respiration of the host, the precise anatomical location and

the presence of local inflammation (Keck, T. et al. 2000, McFadden, E.R., Jr. et al. 1985)

(Figure 2.6). These factors can result in an overall variability of temperature in this

niche ranging from 25.3±2.1°C in the nasal vestibule to 33.9±1.5°C in the nasopharynx,

generally being several degrees below core body temperature (Keck, T. et al. 2000). N.

meningitidis has evolved to rapidly and efficiently adapt its metabolism to even minor

temperature changes in the environment. During the development of invasive disease,

N. meningitidis passes from the lower temperatures in the upper airway to the core

body temperature of 37°C or higher with a febrile response to infection (Cabanac M.

1990). Within the bloodstream although the increased temperature and the abundance

of nutrients promote the fast growth of the bacterium, the presence of the complement

cascade components, immunoglobulins and immune cells represent a big threat,

meanwhile. Therefore a rapid adaptation to the new environment is required, in fact

25

approximately 30% of the genes in the genome are dramatically regulated on entry to

whole human blood (Echenique-Rivera, H. et al. 2011), triggering an immune evasion

response. Key antigens and virulence factors such as capsule biosynthesis (CssA),

sialylation of LPS (Lst) and fHbp involved in immune evasion and in establishing

invasive disease, show increased expression at 37°C relative to lower temperatures

(Loh, E. et al. 2013, Loh, E. et al. 2016). However, the role of lower temperature on the

expression of virulence factors has received considerably less attention. Recently, a

comparative proteomic study showed that 375 proteins were differentially expressed

between 32°C and 37°C (Lappann, M. et al. 2016).

Figure 2.6 Within the host N. meningitidis encounters different niches. Temperature is one of

the key signal sensed by N. meningitidis to determine its environment. The temperatures that N.

meningitidis encounters during transmission, colonization and invasion are reported.

Neisserial Heparing Binding Antigen (NHBA)

NHBA is a surface exposed lipoprotein that is specific to Neisseria species. NHBA is

one of the major antigens of the serogroup B meningococcal vaccine, 4CMenB (Serruto,

D. et al. 2012), and induces antigen-specific bactericidal antibodies in both animals and

humans (Serruto, D. et al. 2010); (Giuliani, M.M. et al. 2010). The nhba gene is

26

ubiquitous in meningococcal strains of all serogroups and it is also found in N.

gonorrhoeae as well as in different commensal neisserial species (Bambini, S. et al. 2009);

(Jacobsson, S. et al. 2006, Muzzi, A. et al. 2013). Analysis of gene sequences from

genetically diverse serogroup B strains revealed the existence of more than 400 distinct

peptides, which are associated with clonal complexes and sequence types

(Comanducci, M. et al. 2002, Jacobsson, S. et al. 2006, Muzzi, A. et al. 2013).

Considerable variation is observed at the level of primary amino acid sequence which

ranges in length from approximately 430 to 500 residues (Figure 2.7). Most variability

is observed at the level of the amino-terminal region, which is annotated as

intrinsically unfolded by commonly used structure prediction algorithms (Vacca, I.

2014). In contrast, the carboxyl-terminal region consists of a single 8-stranded anti-

parallel beta-barrel structure and is highly conserved (Esposito, V. et al. 2011).

The two domains are linked through an arginine-rich motif which is responsible for

NHBA binding to heparin in vitro and contributes to increased survival of the un-

capsulated N. meningitidis in human serum (Esposito, V. et al. 2011, Serruto, D. et al.

2010). Conversely, it has been recently shown that NHBA plays an integral part in

binding heparin sulfate proteoglycans on epithelial cells and thus directly mediates

adhesion of N. meningitidis (Vacca, I. et al. 2016). Furthermore, the presence of this

arginine-rich domain was implicated in DNA-binding and could therefore also play a

role in the formation of neisserial microcolonies and biofilms (Arenas, J. et al. 2013).

NHBA can be processed by the meningococcal protease NalP and human lactoferrin

(hLf). Cleavage occurs either upstream and downstream of the NHBA Arg-rich region

resulting in one of two possible cleavage fragments termed C2 and C1, respectively

(Serruto, D. et al. 2010).

It was also demonstrated that the C-terminal fragment (C2), released upon NalP

proteolysis, alters endothelial cell permeability by inducing the internalization of the

adherens junction protein VE-cadherin, which is in turn responsible for the endothelial

leakage. Thus, the NHBA-derived fragment C2 might contribute to the extensive

vascular leakage typically associated with meningococcal sepsis (Casellato, A. et al.

2014).

27

Figure 2.7 NHBA protein schematic view and variability. NHBA protein sequence reflects a

modular structural organization, where it is possible to recognize three main domains (A, B and

C). The presence of an insertion sequence of 60 amino acids, present only in some of the NHBA

peptides (Insertion IB), allows to discriminate between long or short isoforms. Functional sites

are represented by the Arg-rich region (in brown), by the NalP cleavage site (in green) and by

the human lactoferrin cleavage site (in grey). The C-term of the protein, corresponding to

module C is highly conserved and is represented by a beta-barrel structure. The lower graph

shows the percentage of amino acid conservation between the different peptides along the

protein sequence (adapted from (Vacca, I. 2014)).

The upstream regulatory region of the nhbA gene is characterized by the presence of

the 150-bp Contact Regulatory Element of Neisseria (CREN) in strains, such as MC58,

belonging to clonal complex ST-32. This regulatory element is specific to pathogenic

Neisseria species and is involved in the induction of the downstream associated genes

upon contact with target eukaryotic cells (Deghmane, A.E. et al. 2002). NHBA

expression is known to be induced after incubation of bacteria with epithelial cells in

the CREN-containing strain MC58, while its expression remains unaltered in the

CREN-lacking strain 8013 (Deghmane, A.E. et al. 2003). It was therefore proposed that

cell contact increased NHBA levels on meningococcal surface in the ST-32 invasive

hypervirulent strains and that increased expression of nhba upon host contact might at

least partially account for the hypervirulent phenotype of this clonal complex.

28

3. Results

3.1. NHBA expression and surface exposure are temperature-dependent

NHBA is an important virulence factor for N. meningitidis and it is also protective

antigen able to elicit an immune response in preclinical and clinical trials (Serruto, D. et

al. 2010). Therefore, understanding the mechanisms that drive NHBA regulation is an

important goal to better understand N. meningitidis pathogenesis and vaccine induced

response. We therefore investigated how physiologically relevant temperatures, which

mimic the different stages of pathogenesis, may affect the expression levels of NHBA.

Strains MC58, M11719 and 8047 were grown overnight on GC agar plates at

physiologically relevant temperatures ranging from 28°C up to 40°C. We found that

NHBA expression was thermoregulated in an inverse manner to fHbp, with higher

expression at lower temperature. Western blot analysis showed that NHBA expression

was at its highest level between 28°C and 30°C in all these strains and that its

expression decreased markedly with increasing temperatures (Figure 3.1 A). In

contrast, fHbp expression was highest at elevated temperatures and decreased with

temperature reduction. In order to understand whether temperature regulation of

NHBA was conserved among different N. meningitidis isolates, we expanded our

analysis to a broader panel of strains belonging to different clonal complexes, carrying

different variants and also long or short isoforms of NHBA (Figure 3.1 B and Table 3.1).

NHBA expression levels were variable among the different strains and showed

different processing patterns depending on the strain background and NHBA

variant/isoform present (Table 3.1). Despite different variants and expression levels

between the strains tested, all strains showed increased levels of NHBA at 30°C

compared to 37°C. As NHBA is a surface exposed neisserial protein, we confirmed that

increased expression levels of NHBA also resulted in increased levels of NHBA

exposed on the bacterial cell surface using flow cytometry (Figure 3.1 C).

29

Figure 3.1 NHBA expression and surface exposure are increased at reduced temperatures. (A)

The defined strains were grown overnight on GC agar plates at the indicated temperatures.

Whole cell lysates were prepared and separated by SDS-PAGE prior to Western Blotting. The

indicated proteins were detected using mouse-polyclonal antisera. Hfq served as loading

control between different samples. In MC58 strain the full-length protein migrates with an

apparent molecular weight of approximately 62 kDa, while other bands at approximately 49

kDa results from the bacterial proteases’s processing (Serruto, D. et al. 2010). (B) The defined

strains were grown in GC broth at 30°C or 37°C until OD600 0.25. Whole cell lysates were

prepared and separated by SDS-PAGE prior to Western Blotting. Different strains express

different variants and isoforms of NHBA (See table 3.1 for strains details), however the bands

specificity is confirmed by the nhba deletion mutant (Δnhba) generated for each strain. (C) Flow

cytometric analysis of strain MC58 showing surface exposure of NHBA at the indicated

temperatures.

30

Strain Country of origin

Year of isolation

Capsular group

Clonal complex

NHBA peptide variant

NHBA isoform

M11205 USA 2003 B 41/44 p0001 Long

M11822 USA 2004 B 41/44 p0001 Long

NGH38 Norway 1988 B uaa p0002 Long

MC58 UK 1985 B 32 p0003 Long

M10935 USA 2003 B 35 p0058 Long

M14933 USA 2006 B 32 p0003 Long

M10713 USA 2003 B 41/44 p0010 Short

M03279 USA 1997 B 41/44 p0011 Short

N16/07 Norway 2007 B 41/44 p0029 Long

M11204 USA 2003 B 41/44 p0029 Long

M10282 USA 2003 B 41/44 p0002 Long

M07-0240679 UK 2007 B 269 p0017 Short

M11719 USA 2003 B 162 p0020 Short

M16453 USA 2007 B 41/44 p0144 Short

M18070 USA 2008 B 162 p0020 Short

8047 USA 1978 B 11 p0020 Short

Table 3.1 List of natural strains reported in Figure 3.1. Main characteristics are indicated. For

each of them a nhba deletion mutant was generated. *Unassigned

31

3.2. Mutations and deletions in the 5’UTR and 5’TR of nhba affect expression

The nhba gene was originally annotated as NMB2132 according to its location within

the genome sequence of the strain MC58 (Tettelin, H. et al. 2000). MC58 nhba locus

schematic view (upper panel) and details of the intergenic region and the 5’TR of nhba

(lower panel) are reported in Figure 3.2 A. The conservation of this region among all

the Neisseria species present in the PubMLST database is also reported (mid panel). The

region shows a very good conservation among 8373 Neisseria strains (green and

greeny-brown bars). Red bars indicate the lack of conservation (5.8% strains) which

corresponds to the CREN sequence, specific for ST-32.

Upstream and in the same orientation of nhba is located NMB2133. A Rho-independent

terminator is predicted around 22 nucleotides downstream (Figure 3.2 A lower panel).

qRTPCR experiments confirmed that no co-transcription is detectable among NMB2133

and NMB2132 (data not shown). A putative promoter (Pnhba) was identified upstream

the CREN sequence.

The annotated translational start site is boxed in black and the ribosomal binding site is

also indicated and underlined. However, in frame with the annotated one and just next

to it, two more putative translational start sites were identified (boxed with dashed

lines). Moreover, within the CREN sequence and still in frame with the annotated one,

another putative translational start site (boxed in green) was identified, carrying also

an alternative ribosomal binding site (underlined in green) (Deghmane, A.E. et al.

2003). In order to investigate which one corresponds to the initiation of translation in

correlation with temperature changes, a series of site-directed mutagenesis were

performed (Figure 3.2 B). As shown by Western blot analysis, small deletions or single

base mutations affecting these sites (Mut_1-4) led to decreased or abolished NHBA

expression, without affecting thermoregulation (Figure 3.2 C left panel). We identified

a T-rich region in the 5’TR of nhba, 20 nucleotides downstream to the putative

translational start site. In silico secondary structure prediction of 5’UTR+50bp in the

coding sequence suggested a direct interaction between the T-rich region and the

ribosomal binding site (data not shown). Synonymous mutations in this region (Mut_5-

7) led to an overall decreased expression of NHBA, without affecting NHBA

thermoregulation (Figure 3.2 C right panel).

32

A

B

C

33

Figure 3.2 Schematic representation of nhba locus in MC58 strain and site-directed

mutagenesis. (A) Schematic representation of the nhba locus (upper panel). The conservation of

the indicated intergenic region obtained by multiple sequence alignments of 8373 Neisseria

strains present in PubMLST database is shown (mid panel). Each bar represents the percentage

of conservation of the corresponding nucleotide. Green bar = 100% identity; Greeny-brown bar

< 100% identity; Red bar < 30% identity. The nucleotide sequence of the indicated region is

reported (lower panel): the 3’ region of NMB2133 is boxed in dark grey and the 5’ region of nhba

(NMB2132) is boxed in light grey. The nucleotides pairing in the stem region of the Rho-

independent terminator are underlined (dot line). In the intergenic region downstream of

NMB2133, a putative promoter sequence was identified (Pnhba). The -35 and -10 elements of the

Pnhba are indicated and the putative transcriptional start site is indicated and highlighted in

bold. The contact regulatory element of Neisseria (CREN), a 150-bp sequence specific for

pathogenic Neisseria species, is present in MC58 strain immediately upstream of the ribosomal-

binding site (Deghmane, A.E. et al. 2003) and is boxed in green. The ribosome binding sites are

underlined and the translation start sites are boxed. (B) Schematic representation of site-

directed mutagenesis. Red dashes indicate nucleotides deletion, red nucleotides indicate non-

synonymous mutations (Mut_3 and Mut_4) and synonymous mutations (Mut_5-Mut_7). (C)

Mutant strains were grown overnight on GC agar plates at the indicated temperatures. Whole

cell lysates were prepared and separated by SDS-PAGE prior to Western Blotting. The indicated

proteins were detected using mouse-polyclonal antisera. The *a symbol indicates a non-specific

band used as loading control between different samples.

34

3.3. NHBA is expressed during the active growth

Previous analysis only represented a single time point of NHBA expression levels

during mid-exponential growth. In order to determine how NHBA expression

progresses during the entire growth of N. meningitidis, we grew NGH38 strain at either

30°C or 37°C in 50 ml of liquid culture. We took samples for RNA extraction and

Western Blotting at various stages during the entire growth curve of the strain (Figure

3.3). By culturing the strain in flasks, we were able to take samples every hour and

follow the growth for ten hours (Figure 3.3 A). Both curves reached high OD600 values

however, under these experimental conditions, bacteria grown at 30°C showed longer

lag phase (T0-T4) and slightly lower OD600 values at the end of the growth, respect to

those grown at 37°C. Firstly, we investigated the transcriptional profile of nhba, fHbp

and adk during growth in liquid culture (Figure 3.3 B). We determined by qRT-PCR

that nhba transcript was most abundant during active bacterial replication. Once

bacteria entered stationary phase, transcription of nhba was almost abolished. During

the active replication, nhba RNA steady state levels were higher at 30°C respect to 37°C.

Conversely, fhbp transcript resulted to be slightly more abundant at 37°C respect to

30°C, while no differences were observed for adk comparing the two temperatures. The

nhba RNA steady state level expression profiles during the entire growth were

confirmed also at protein level by Western blotting, showing that NHBA is most

abundant during the active growth at both temperatures (Figure 3.3 C).The higher

expression of NHBA also resulted in higher processing at 30°C.

35

36

Figure 3.3 NGH38 growth profiles and NHBA expression levels during the entire growth. (A)

Growth profiles of NGH38 strain in 50 ml of MCDMI liquid medium at 30°C (blu line) or 37°C

(red line). Bacteria were grown for 10 hours and RNA isolation and samples for Western blot

analysis were collected every hour. (B) nhba, fHbp and adk RNA steady state levels were

quantified by qRT-PCR and relative expression levels were determined normalizing to 16S-

rRNA. (C) Western blot analysis of NHBA on whole cell lysates collected at the indicated time

points. In NGH38 strain the full-length protein migrate with an apparent molecular weight of

approximately 62 kDa, while other bands at approximately 49 kDa results from the bacterial

proteases’s processing (Serruto, D. et al. 2010). The *a symbol indicates a non-specific band used

as loading control.

Therefore we decided to compare more in details the expression of NHBA in response

to temperature by using standard in vitro conditions. We grew MC58 strain at either

30°C or 37°C in 7 ml liquid culture and took samples for RNA extraction and Western

Blotting at various stages during the entire growth curve of the strain (Figure 3.4 A).

Firstly, we investigated the transcriptional profile of nhba during growth in liquid

culture (Figure 3.4 B). By qRT-PCR we confirmed that nhba transcript was most

abundant during active bacterial replication. Once bacteria entered stationary phase,

transcription of nhba was almost abolished (Figure 3.4 B). Although there was a trend

for slightly increased nhba transcript levels at 30°C, these transcriptional differences

were only significant in late exponential phase. We therefore next examined the NHBA

protein expression profile at the same growth phase as the nhba transcript levels. In

accordance with the data for the nhba transcript, we observed that NHBA level at each

temperature were highest during exponential growth of the bacterium (Figure 3.4 C).

As seen previously, NHBA protein levels in each growth phase were always higher at

30°C relative to 37°C (Figure 3.4 C). However, although nhba transcript is barely

detectable at stationary phase with no differences between the two temperatures,

becomes evident that the protein is still abundant in bacteria cultured at 30°C. To

quantify the differences in protein expression levels between the two temperatures

tested, we used relative protein quantification (Figure 3.4 D). This analysis determined

that the amount of NHBA at 30°C was approximately 3-5-fold higher than at 37°C.

37

Figure 3.4 NHBA is expressed during the exponential phase and its expression at 30°C is

higher relative to 37°C at both RNA and protein levels. (A) Growth profiles of MC58 strain in

GC liquid medium at 30°C (continuous line) or 37°C (dot line). Samples for western blot

analysis and RNA isolation were collected at early exponential phase (OD600 ̴0.3), late

exponential phase (OD600 ̴0.9) and at stationary phase (OD600 ̴1.1), as indicated by the star

symbols. (B) nhba RNA steady state levels were quantified by qRT-PCR and relative expression

levels were determined normalizing to 16S-rRNA. (C) Western blot analysis of NHBA on whole

cell lysates collected at the indicated time points. (D) Relative protein quantification performed

with ImageJ 1.6 software. In MC58 strain the full-length protein migrate with an apparent

molecular weight of approximately 62 kDa, while other bands at approximately 49 kDa results

from the bacterial proteases’s processing (Serruto, D. et al. 2010). The *a symbol indicates a non-

specific band used as loading control and for relative protein quantification. All the data

represent the mean +/- SEM from three independent biological replicates and were analyzed by

Two-way Anova followed by uncorrected Fisher’s LSD multiple comparison test (**** p

38

3.4. NHBA thermoregulation is not driven by the nhba promoter

To investigate the molecular mechanisms involved in NHBA thermoregulation a nhba

deletion mutant (Δnhba) was generated in the MC58 strain background by replacement

with an erythromycin antibiotic resistance cassette and different isogenic

complementation mutants were generated. To test if the genomic context played a role

in nhba regulation, the complete sequence, comprising the entire gene and the

intergenic regulatory region, was inserted into the NMB1428-NMB1429 genomic locus,

generating the Δnhba-C_nhba strain (Figure 3.5 A). As shown by Western blot analysis

and qRT-PCR (Figure 3.6 A), the wild type and Δnhba-C_nhba strains showed the same

expression and thermoregulation of nhba as the wild-type strain, indicating that

placing the wild type sequence of nhba in another genomic locus does not affect nhba

regulation in response to temperature changes at both RNA and protein level.

In order to determine whether the nhba promoter was required for thermoregulation,

we replaced the MC58 wild-type sequence with an IPTG-inducible Ptac promoter,

immediately upstream of the CREN sequence (Figure 3.5 B), generating the Δnhba-

Ptac_nhba strain. We observed an IPTG dose-dependent increase of nhba expression at

both RNA and protein level. However, using the same amount of IPTG for induction,

we no longer observed any differences between the transcript levels at the two

temperatures (Figure 3.6 B). Interestingly, we still observed clear thermoregulation of

NHBA at protein level, albeit less pronounced than in the wild type strain.

To determine whether regulatory elements in the nhba upstream intergenic region

contributed to thermoregulation, we fused the full intergenic region comprising the

promoter, the CREN sequence and the initial part of the coding sequence

corresponding to the first 14 amino acids to a mCherry reporter gene (Figure 3.5 C),

generating the Δnhba-Pwt_mCherry strain. As independent control, we also generated

Δnhba-Ptac_mCherry strain, a reporter fusion under the control of the IPTG-inducible

promoter (Figure 3.5 D). We then analyzed the expression of the reporter gene product

by qRT-PCR and Western blotting of samples collected at either 30°C or 37°C (Figure

3.6 C). The IPTG inducible promoter drove higher transcription of the downstream

gene with respect to Pnhba promoter. However, no significant differences were

observed at protein level. We observed no thermoregulation in either reporter fusion

construct, both at RNA and at protein level, suggesting that the promoter on its own

cannot account for the observed differences in expression at the two temperatures. All

39

together these results highlight that NHBA thermoregulation is at the post-

transcriptional level.

Figure 3.5 Schematic representation of nhba mutants generated by ex-locus complementation.

(A-D) In the MC58Δnhba strain background different mutants were generated by

complementation in the NMB1428-1429 locus: KanR-kanamycin resistance cassette; CmR-

chloramphenicol resistance cassette; LacI – LacI repressor gene; Ptac – IPTG inducible promoter;

mCherry – mCherry reporter gene.

40

Figure 3.6 NHBA

thermoregulation is at

post-transcriptional

level. (A-C) Wild type

and recombinant strains

were grown in GC liquid

medium at 30°C or 37°C,

with the indicated

concentration of IPTG,

where needed. NHBA

and mCherry protein

expression were assessed

by western blotting by

using polyclonal mouse

antisera and monoclonal

mouse antibody

(ab167453, abcam),

respectively (upper

panel). nhba and mCherry

RNA steady state levels

were quantified by qRT-

PCR and relative

expression levels were

determined normalizing

to 16S-rRNA (lower

panel).

41

3.5. Temperature affects nhba RNA half-life

Transcription produces single-stranded RNA molecules that easily form intra- or

intermolecular partially double-stranded RNAs, or associate with proteins, which may

be used to regulate gene expression. To prevent or resolve kinetically trapped

structures, cells use RNA chaperones that in most cases bind RNA non-specifically to

help refold RNA or RNA–protein (RNP) complexes in ATP-independent or ATP-

dependent reactions (Herschlag, D. 1995); (Mohr, S. et al. 2002); (Rajkowitsch, L. et al.

2007). The kinetic problem of RNA folding is dramatically aggravated at low

temperatures. Few specific molecular mechanisms have been described to be involved

in the cold shock response. CspA is the major regulator of the cold-shock response in

E.coli, where the entire process of cold response is well documented. It is a RNA

chaperone, able to bind both DNA and RNA, and its own regulation in response to

temperature downshift is driven by a post-transcriptional mechanism which involves

the 5’UTR region, resulting in higher transcript stability at low temperatures. The N.

meningitidis homologue for cspA is NMB0838 and it resulted to be upregulated at 32°C

in microarray analysis (data not shown). Another class of RNA binding proteins that

accelerate structural rearrangements of RNA, particularly in the cold, are the DEAD-

box RNA helicases. By using ATP as substrate, they mediate RNA conformational

changes, otherwise kinetically unstable (Iost, I. et al. 2013). NMB1368 is a RNA helicase

H present in N. meningitidis. Therefore, we decided to generate deletion mutants of

NMB1368 and NMB0838 in MC58 strain background to investigate if this could have

an effect on NHBA thermoregulation. Mutant strains did not show alterations in

thermoregulation, both by qRT-PCR and Western blotting (Figure 3.7), even if a

slightly decrease in expression was observed in ΔNMB0838 mutant grown at 30°C.

42

Figure 3.7 Deletion of NMB1368 or NMB0838 genes did not affect NHBA thermoregulation.

NMB1368 or NMB0838 genes were deleted in the MC58 wt strain background. Mutant strains

were grown in GC broth at 30°C or 37°C until OD600 0.5. Whole cell lysates and samples for

RNA isolation were collected. (A) Whole cell lysates were prepared and separated by SDS-

PAGE prior to Western Blotting. The *a symbol indicates a non-specific band used as loading

control and for relative protein quantification. (B) nhba RNA steady state levels were quantified

by qRT-PCR and relative expression levels were determined normalizing to 16S-rRNA.

43

We evaluated the nhba RNA decay after stopping active transcription by addition of

rifampicin. The relative RNA amount was quantified by qRT-PCR (Figure 3.8). To

obtain higher accuracy and reliability we used NGH38 as wild-type strain as overall

nhba expression levels were higher compared to MC58. The transcript of nhba at 37°C

showed a very short half-life and was rapidly degraded below the limit of detection

(Figure 3.8). However, nhba transcript decay was found to be directly influenced by

temperature, showing a shorter half-life at 37°C relative to 30°C

Figure 3.8 The nhba transcript has a shortened half-life at elevated temperatures. Strain

NGH38 was grown in GC broth until OD600 0.5 at the defined temperatures. RNA extracts were

prepared at different time points after active transcription was stopped by adding rifampicin.

nhba and 16s RNA abundance were measured by qRT-PCR and quantified relatively to the

levels observed at the start of the experiment. Relative RNA quantity was calculated as 2-(Ct1-Ct0)

and transformed as Y= log(Y). A linear regression was performed for each dataset and plotted

as continuous lines. Data represent the mean from three independent biological replicates ± SD.

44

3.6. NHBA protein shows higher stability at 30°C respect to 37°C

Having determined that NHBA protein levels are strongly affected post-

transcriptionally, we wanted to investigate whether altered protein stability at different

temperatures could account for the observed differences. We therefore grew the MC58

wild type strain in GC broth at 30°C and 37°C to mid-exponential phase and stopped

protein translation by adding spectinomycin. Each culture was then split and

incubated at both 30°C and 37°C and samples for whole cell protein extraction were

withdrawn at different time-points of treatment. NHBA accumulation upon protein

translation arrest was analyzed by Western blots with representative results shown in

Figure 3.9. At 30°C, the amount of NHBA full length protein appears the same for 45-

60 min treatment (Panel A, upper part), whereas treatment at 37 °C show no change for

about 10 min treatment (Panel A, lower part).

Cells grown at 37 °C and treated with spectinomycin at 30 °C show a similar amount

of NHBA for about 20 min treatment (Panel B, upper part), while treatment at 37 °C

shows no changes for 10 min (Panel B, lower part). Overall, these data suggest that

NHBA turnover is higher at 37°C and, accordingly, it appears stable at 30 °C. Thus,

NHBA thermoregulation is additionally exerted by post-translational mechanisms.

Figure 3.9 NHBA protein turnover is directly affected by temperature changes. (A-B) MC58

wild type strain was grown in GC broth until OD600 0.5 at 30°C and 37°C. Whole cell extracts

were prepared at different timepoints after the active translation was stopped by adding

spectinomycin. Protein samples were separated by SDS-PAGE prior to Western Blotting using a

anti-NHBA mouse polyclonal serum. The full length protein is shown. The *a symbol indicates

a non-specific band used as loading control.

45

3.7. NHBA expression levels correlate with susceptibility to complement-

mediated killing by anti-NHBA antibodies

NHBA is one of the three major components of the 4CMen vaccine against serogroup B

meningococcus and is a protective antigen that is able to elicit a robust immune

response. Although NHBA is present in all neisserial species, we have shown that its

expression is variable among strains and moreover is affected by several factors such

as like growth phase and temperature changes. It is therefore paramount to

understand how different NHBA expression levels, either through strain variation or

triggered by different growth conditions affect the bacterium’s susceptibility to killing

mediated by anti-NHBA antibodies.

We therefore investigated whether the different NHBA expression levels at different

temperatures could affect bacterial susceptibility to complement-mediated killing. To

address this hypothesis, we grew strain MC58 at both 30°C and 37°C and confirmed

increased NHBA and decreased fHbp expression levels at 30°C compared to 37°C

(Figure 3.10 A). Functional antigen should be exposed on the bacterial cell surface and

we verified that altered antigen expression levels translated into altered surface

exposure looking at NHBA, fHbp and cps surface exposure by flow cytometry (Figure

3.10 B). Given these observations, we assessed the ability of immune serum raised

against NHBA to kill N. meningitidis grown at the two different temperatures by

determining their serum bactericidal titer (Figure 3.10 C). This assay is also called

serum bactericidal assay (SBA) and was performed at both 30°C and 37°C. While clear

differences in antigen expression were evident by Western blotting and flow cytometry

analysis and trends of the SBA assays reflected these differences, we were unable to

gain statistical significance due to the intrinsic variability of the experiment. We

reasoned that altering the temperature would affect many different processes in the

bacterial cell and confounding pleotropic effects would make it difficult to determine

the precise impact of NHBA expression levels on serum bactericidal killing. For

example, important virulence factors such as fHbp and the neisserial capsule also

respond to temperature changes, but in the opposite way compared to NHBA (Figure

3.10 B). These pleiotropic effects could then mask the role of NHBA in this assay

making it impossible to extract how NHBA expression levels affect bacterial killing in

this assay.

46

Figure 3.10 Temperature driven expression, surface exposure and susceptibility to

complement-mediated killing. The MC58 wild type strain was grown at 30°C or 37°C in MH

broth +0.25% glucose until OD600 0.25 was reached. Bacteria were collected and expression

levels of NHBA and fHbp were determined by (A) Western blotting and surface exposure of the

defined antigens was confirmed by (B) flow cytometry using polyclonal antisera. (C) Serum

bactericidal titers were determined using baby rabbit complement as source of complement

factors (rabbit SBA, rSBA). rSBA titers indicate the dilution of the α-NHBA mouse polyclonal

serum, α-fHbp mouse polyclonal serum or α-cps mouse monoclonal antibody at which 50% of

killing was reached. All the data are representative of three independent biological replicates.

47

In order to circumvent pleiotropic effect on bacterial expression of virulence factors

other than NHBA, we generated a recombinant strain in which NHBA expression was

under the control of an IPTG-inducible promoter (MC58 Δnhba-Ptac_nhba). This assay

allowed us to perform the experiment under identical conditions while varying NHBA

expression through addition of different IPTG concentrations. We confirmed protein

expression and surface exposure of NHBA in these cultures using Western blotting and

flow cytometry analysis (Figure 3.11 A and B, respectively).This then enabled us to

extract the role of NHBA expression levels in the ability of anti-NHBA antiserum from

mice to mediate complement-dependent killing through the rSBA (Figure 3.11 C). We

observed that rSBA titers correlated directly with NHBA expression levels, as

confirmed by statistical analysis.

48

Figure 3.11 Correlation between NHBA expression, surface exposure and susceptibility to complement-mediated killing by anti-NHBA antibodies. MC58

Δnhba, wild type and Δnhba-Ptac_nhba strains were grown in MH broth +0.25% glucose until OD600 0.25 at 30°C or 37°C as indicated. IPTG was added, where

needed, at the indicated final concentrations. Bacteria were collected to perform (A) Western blotting, (B) flow cytometry analysis and (C) serum bactericidal assay.

The relative quantification obtained by (A) densitometry analysis and (B) FACS geometric mean calculation were estimated. All the data represent the mean +/- SD

from three independent biological replicates and a linear regression was performed on each dataset (black lines indicate the linear fit, dot lines indicate the 95% CI).

A Pearson correlation test was used to assess the goodness of correlation between the three different datasets (Densitometry/SBA titers: Pearson r = 0,962; P=0.009.

49

Densitometry/FACS GeoMean: Pearson r = 0,972; P=0.006; FACS GeoMean/SBA titers: Pearson r = 0,925; P=0.008). Western blotting relative quantification of 0.050

mM IPTG samples were not taken into account for the linear regression as these were found to be out of the linearity range of the assay

50

3.8. NHBA regulation during invasion

It has been previously shown that NHBA can be processed by the meningococcal

protease NalP and human lactoferrin. Cleavage occurs either upstream and

downstream of the NHBA Arg-rich region resulting in one of two possible cleavage

fragments termed C2 and C1, respectively (Serruto, D. et al. 2010). To mimic the

invasive condition MC58 and NGH38 wild type strains were grown until early

exponential phase and then incubated at 37°C in presence of 25% of human serum

during a time course, up to 2 hours (Figure 3.12 A and B). After 15 minutes of serum

incubation NHBA was induced and strongly processed in both strains. In fact, despite

the appearance of the lower bands typical of the N terminal portion of NHBA after

cleavage of the protein, the band corresponding to the full length protein appeared to

retain the same intensity or even more suggesting induction of the full length NHBA as

well as concomitant processing. Within the blood there is an abundance of components

with proteolytic activity, so it is not surprising to see an increase in cleavage. The